You are on page 1of 17

Hoards from the Neolithic

to the Metal Ages
Technical and codified practices
Session of the XIth Annual Meeting of the
European Association of Archaeologists

Edited by

Caroline Hamon
Benedicte Quilliec

BAR International Series 1758

co.This title published by Archaeopress Publishers of British Archaeological Reports Gordon House 276 Banbury Road Oxford OX2 7ED England .com The current BAR catalogue with details of all titles in print. Session of the XIth Annual Meeting of the European Association of Archaeologists © the individual authors 2008 ISBN 978 1 4073 0197 6 Printed in England by Butler and Tanner All BAR titles are available from: Hadrian Books Ltd 122 Banbury Road Oxford OX2 7BP England bar@hadrianbooks. prices and means of payment is available free from Hadrian Books or may be downloaded from BAR S1758 Hoards from the Neolithic to the Metal Ages: Technical and codified practices.archaeopress.

Such axes were believed to represent hidden trade-goods. Fig. new studies focusing on the nature and interpretation of the Dutch Neolithic depositions remained absent. zijn vooral gevonden in moerassige plekken. However until now. These axes associated with the Funnel Beaker Culture are often retrieved from waterlogged places either single or as part of multiple object hoards. micro-wear and spatial analysis it will be demonstrated that these objects are remnants of a highly structured phenomenon that was practiced throughout the research area.). Wentink K. three clusters of flint at Hoge Vaart (Hogestijn & Peeters 2001: 41). welke toegeschreven worden aan de Trechterbeker cultuur. when all ‘profane’ explanations could be excluded. zowel alleen of als onderdeel van een meervoudig depot. Door middel van een metrische-. The earliest intentional depositions known in the Netherlands were dated to the Late Mesolithic.C. By means of a metrical. & A. The Dutch TRB (part of the TRB-Westgroup) has the advantage of a more or less restricted geographical distribution within the Netherlands. van Gijn . These concerned pottery vessels that were buried in pits together with pieces of antler. B. left there by merchants to be retrieved later. The fact that most were found in waterlogged places in particular formed the basis of speculation as to the nature of these objects.3400-2850 cal. Deze bijlen. are widely dismissed (Fontijn 2002: 19). these are still a subject of debate (pottery vessel and red-deer antler from Bronneger (Louwe Kooijmans 2001: 112). 29-43 NEOLITHIC DEPOSITIONS IN THE NORTHERN NETHERLANDS Karsten WENTINK and Annelou van GIJN Abstract: The aim of the present paper is to explore the nature of the Dutch Middle Neolithic flint axe depositions and interpret these on a cultural level.K. It was not until the Middle Neolithic that depositional practices became more structured and common. The focus of this paper will be the depositional practices associated with the Middle-Neolithic Funnel Beaker Culture (TRB . Presently such interpretations. Although some other Early Neolithic finds are known that could be interpreted as intentional deposition. so clearly devised by minds influenced by western capitalism. Samenvatting: Het doel van het huidige artikel is een overzicht te geven van de Nederlandse midden-neolithische deposities en deze te interpreteren op een cultureel niveau. gebruikssporen-. Or perhaps they were treasures hidden in times of trouble. Research area (black square).Neolithic Depositions in the Northern Netherlands. being mostly confined 29 . 1. bone and wood (Louwe Kooijmans 2001: 512). p. en ruimtelijk-analyse zal worden aangetoond dat deze objecten onderdeel uitmaken van een uiterst gestructureerd fenomeen dat ten uitvoer werd gebracht in het gehele onderzoeksgebied. A ritual explanation was only proposed. Surely people would not have been living in such inhospitable areas. Introduction Already in the 19th century discoveries of groups of large axes puzzled those confronted with them.

1) located in the province of Drenthe (Bakker 1982). lost. Do these remarkable finds from waterlogged places indeed reflect prehistoric ritual behaviour. interpretations like 30 . are found in numerous depositions containing either single or multiple objects that were retrieved from waterlogged places. The most dense concentration of TRB finds however. This till plateau was formed in the Saalian ice-age.Hamon C. stolen or destroyed. pottery vessels (probably containing foodstuffs) and disc-wheels. Ter Wal has convincingly argued for the existence of single object hoards containing only one large axe deliberately placed in the peat (Ter Wal 1996). they are rectangular in cross-section. it rarely contains good quality flint. and on one occasion the complete hoard was lost and is only known from 19th century written sources (see Pleyte 1882: 52).Hoards from the Neolithic to the Metal Ages : technical and codified practices to the northern half of the country. are found on the Drenthe Plateau (fig. such as horns of cattle. Although several multiple object hoards consist of only 2 axes (n=7) most contain 3-5 axes (n=9) with only a few containing more. which were mainly produced in Northern Germany and Denmark. Furthermore. votive hoards. flint nodules and other tools (Achterop 1960. The latter however don’t only consist of axes but also of flint nodules. general patterns and interpretational framework At present 20 multiple object hoards are known from the Netherlands containing multiple axes. which patterns can be observed and how should these be interpreted? What was the life-history of these objects and how can these be linked to the lives of either individuals or groups? The present paper presents preliminary results of the Research Master thesis of the first author (Wentink 2006). workshop hoards. This caused TRB people to be dependent on exchange contacts to acquire good quality flint axes. The imported northern TRB axes. If so. the first author. roughouts. making them easily distinguishable from the oval axes made in the Atlantic tradition. the remaining three were unfortunately not of a distinguishable character (Achterop 1960. The fact that all known hoards have been found during the last 50 years of this reclamation one can only imagine what has been lost. as well as the spatial and metrical analysis was performed by One of the problems in dealing with depositions and rituals is determining when a find should be considered as such. and forms part of the second author’s research project “The social significance of flint for Neolithic and Bronze Age communities”(Van Gijn in prep. The latter are predominantly found in the southern half of the Netherlands. the Dutch TRB finds from the large bogs are basically confined to peat track ways and finds of an alleged ritual character. Although evidence of earlier TRB activity in Northern Germany concentrates in the wetlands formerly exploited by the Mesolithic predecessors (Midgley 1992: 311). The reclamation of peat began as early as the 17th century and continued well into the mid 20th century. contextual information is often of poor quality or completely lacking. Nine hoards. the depositing of disc-wheels is exclusively dated to the Single Grave Culture (SGC) (Van der Waals 1964). It is the aim of the study to investigate and interpret this phenomenon of deposition on a cultural level. they are completely absent from hoards (Bakker 1982: 95). long blades or other flint tools. Although many axes ended up in museum collections. TRB flint axes share a very distinct technological feature. Several objects were left in the field. however. researchers distinguished between several different categories of finds. can be placed in the TRB period and eight can be attributed to the subsequent SGC. eds. As of yet no evidence is present to suggest that the bogs were actively exploited as part of the subsistence strategy by the TRB. 2008 . The compilation of the database. Ter Wal 1996). unretrieved traders stock. In the Holocene period it was for the greater part surrounded by large peat areas of which the ‘Bourtanger Bog’ is one of the biggest in Europe. several other types of objects were placed in the peat in Neolithic times. Although some of these southern axes did reach the Plateau and were found in graves. The Dutch hoards. Although the till contains many large boulders that were used for the construction of the passage graves. whereas the microwear analysis was done in cooperation. For a long time. Ter Wal 1996). (Schuhmacher 1914). based on typology. There is no evidence to suggest that high quality flint axes were locally produced (Beuker 2005: 277). etc. & Quilliec B. Finds could be interpreted as hidden treasure. Although the former occurred during the TRB period. Most settlements and megalithic tombs were located on the Drenthe Plateau. However. Many of the hoards were discovered during peat-cutting activities at the end of the 19th and beginning of the 20th century.).

For the analysis a stereo-microscope (magnifications 10-160x) and an incident light microscope (magnifications 100-500x) were used. Micro-wear analysis From the recorded objects 69 axes were selected for micro-wear and residue analysis.). These patterns might be reflected by the context in which the objects are found. In order to interpret selective deposition one should not just look at a single hoard or object. Stout 2002. thus enabling highpower functional analysis on large objects. & A. However. Sources & Methodology Database The main tool used to gain access to patterns and subsequent interpretations of depositional practices was the compilation of a database (MS Access) containing information on hoards. This applies not only to depots. Many of the older finds. they contained no complete flint axes. From the latter. How do we determine to which category our finds belong and more importantly. With the aid of the stereo-microscope a general survey of the object was carried out and obvious traces of residue located. which mainly reflect modern western logic. The incident light microscope used was attached to an adjustable stand. Site information was retrieved from literary sources and the Dutch National Archaeological Database (Archis) and included amongst others all known megalithic monuments. Numerous sources were used for the compilation of the database. in prep. single finds and sites related to the TRB culture. Wentink K. namely deposition in a certain context and the lack thereof in others. Moreover. Leiden University. excavated settlements and many find-scatters. profane interpretation. this approach should be equally suitable for Neolithic depositions.Neolithic Depositions in the Northern Netherlands. Interpretation should therefore be based on these patterns and not on individual objects. but also for that matter. among which objects from multiple object hoards. Hampton 1999.K. p. axes were examined and contextual and metrical information was recorded. finds from megalithic tombs and a collection of stray finds. The axes of several multiple object hoards in the collection of the Drents Museum were part of a travelling exhibition and were therefore not available for this research. peat trackways. Comparison of recorded 31 . were recovered. having had the same treatment before deposition or showing wear-traces of specific activities. stone cists. or the lack thereof. only some axe fragments (Van Gijn. The definition of a site here being a location where multiple finds. Although Fontijn’s research deals with Bronze Age depositions. polish. Axes from a variety of contexts were examined. The selected sample has been subjected to residue and use-wear analysis at the Laboratory for Artefact Studies. For examination with the incident light microscope some objects were partially cleaned with alcohol to remove finger grease after the absence of potential residue was attested using both microscopes. Sites or objects from the above sources were only incorporated when the find-location could be pinpointed with at least 2 km. Photographs were taken with a Nikon DXM1200 digital camera. the use-life of these objects may display patterning such as originating from the same source. striations and residues were recorded (Van Gijn 1990). If these finds reflect prehistoric ritual behaviour. TRB flat-graves. 1038 of which describe individual axes. for which only a rather vague description of the find-location was available were therefore ignored. The question whether or not a specific object is a ritual deposition therefore becomes less relevant and more dependent upon the entire material context from that place and period. Object information was partly retrieved from literary sources and partly from museum collections. but at all the evidence and at the patterns this evidence brings to light. thus excluding multiple object hoards. Phenomena such as edge-removals. Fontijn 2002). rounding. supposedly single object hoards. 29-43 these present us with a number of problems. Before a find could be labeled as being ‘ritual’ one had first to dismiss any ‘logical’. van Gijn . do these categories. conform to prehistoric ones? The latter question in particular should be answered negatively. Presently the database contains 1645 records. there should be consistencies that are shared by all finds which are remnants of that ritual. not related to a single act. When dealing with deposition Fontijn therefore proposes to abandon these outdated attempts to align prehistoric behaviour with our own categories and to look instead at patterns within aspects of the objects themselves and the context in which they are found (Fontijn 2002). Two excavated TRB settlements were included. many anthropological sources have shown that profane and ritual are two things that are often intertwined and that no real distinction can be made between the two (Hermkens 2005. to any other type of archaeological data.

to peak at the 10032 . 100-1000 m or more than 1 km. making it possible to plot them onto a series of maps using the GIS software MapInfo (Version 7. The number of discarded flint axes. 2. The aim of the residue and micro-wear analysis was to obtain information on the use of flint axes in general and also to gain information on the use-life of individual axes. & Quilliec B. as can be seen in figure 2. both stone and flint axes are similar in terms of length. with a peak in the 75-100 mm range. although stone axes tend to be a bit larger with the exception of a few extremely large flint axes. 2008 . The latter could only be used to give an approximate overview of find distributions. it is clear that these axes are often found in. Frequencies of length relative to raw material. something that is also observed in Denmark (Tilley 1996: 101). eds. The aim of the analysis was to obtain information on the relation between sites and finds and also to investigate their relation to the landscape. The records were combined with cartographical information including geological maps. It should be noted that the observed length of the axes represents the end-stage of the ‘life’ of each individual axe (marginalizing lost axes). At present the database contains records of 1038 axes. while axes from wet or border contexts are larger than average. on or near the underlying sand.Hoards from the Neolithic to the Metal Ages : technical and codified practices phenomena with experimentally used tools led to the interpretation of the object’s functional life. making the contemporary land-peat border unrepresentative of the Neolithic situation. Axes from dry contexts are generally smaller than the average 128 mm. Ter Wal carried out an extensive metrical analysis on a sample of 433 axes from the Drents Museum in Assen and concluded that it appears that axes from a wet context are generally much larger than those from a dry context (Ter Wal 1996). This suggests that at the time they were deposited at the edge of the peat. In the cases where the vertical position is mentioned. historical maps. The older find descriptions in particular do not contain detailed metrical information about the axe. shows a gradual increase as length decreases. the patterns are similar to Ter Wal’s observations. Also a number of axes had to be dismissed from the analysis due to their being incomplete. Flint axes from border contexts (i. which would subsequently be covered by a layer of peat up to several meters thick. soil maps. Depositing an object at the edge of the peat would mean that the axes would be positioned near the underlying sand.e. Spatial analysis For each object or site in the database coordinates (Netherlands National System) were recorded. For the variable ‘length’ 789 axes could be used in the analysis which is 76% of the total number of axes in the database and can therefore be seen as a representative number. Of many of the objects in the database only an approximate find location was known. The problem here is that many older finds are often only described as coming from the peat whereas their vertical position relative to the underlying sand is not mentioned.Hamon C. Fig. Metrical analysis General observations The most conventional way of studying stone or flint axes is by means of a metrical analysis. As can be seen in table 1. Although the current dataset is much larger. It is precisely this context from where most multiple object hoards seem to originate. Stone axes however show a more extreme pattern of discarding which already starts in the 125-150 mm range. describing the accuracy of the record. This could vary from an accuracy in the range of 1-10 m. However axes of different lengths they are not evenly distributed over the landscape. land-use maps and a detailed digital elevation model of the province of Drenthe (AHN). transition zones from dry to wet places) are almost double the length of the average. 10-100 m.0). This vertical position is of interest as the peat gradually grew over time. and in the light of the nature of this research it would take too much time to measure all available axes manually. these however cannot all be used in a metrical analysis. For this reason an additional variable was added to each set of coordinates.

Stone axes have a blunter edge. & A.9% come from a grave context. Virtually all axes are below the 150 mm line which according to Bakker separates imported axes from locally produced axes (Bakker 1979). 29-43 Table 1: Number and average length of stone and flint axes from varying geological contexts. we find they are predominantly large (>150 mm). For this pattern a functional explanation can be offered. With flint axes it is predominantly the sharp edge which makes the tool functional. As they were produced from poor quality flint. The smaller . but their raw material is less susceptible to damage. 125 mm range. making it more suitable for fine carpentry. resulting in an overall pattern of slightly bigger discarded stone axes relative to the flint axes whose effectiveness is less linked with tool length or weight.5% of the axes smaller than 150 mm come from graves. This. Bakker has suggested 150 Fig.K. p. While a fair number of flint axes are still used into the 50-75 mm range. TRB axes The dataset used above reflects a palimpsest situation as no distinction has been made between axes belonging to different cultures or periods. although of poor quality. van Gijn . In fact this is quite difficult to do since many axes cannot be attributed to a specific Neolithic culture apart from the larger axes . with the aid of experiments. There are however clear patterns within this dataset that are not influenced by the over-representation of these groups.9% of the total dataset. When examining the individual axes from grave contexts. . Reduced length and therefore weight of the stone axe has a negative effect upon its effectiveness as a tool. It seems that stone axes are discarded slightly bigger than flint axes. 3. If we only consider axes attributed to the TRB. When plotting the relative distribution of axes from grave contexts per length group this becomes particularly clear as can be seen in figure 4. From the axes smaller than 150 mm 25. This local flint was moreover transported by the Saalian glaciers from Denmark and Northern Germany to the Netherlands. that the main difference in usage between flint and stone axes lies in the configuration of their edges and inherent qualities of the raw material (Olausson 1983). A flint axe blade is generally sharper and thinner resulting in a deeper penetration with each blow. mm being the maximum length of locally produced axes (Bakker 1979). the raw material defined to a greater degree the eventual shape of the artefact.locally produced . stone axes in that range are virtually absent. imported axes (52. it 33 . This means that among the axes attributed to the TRB culture both axes from graves and the large imported axes are over-represented. Based on the nature of the local raw material.axes do not have any defining characteristics.which can often be assigned to either the TRB or SGC. together with the total weight of the tool makes it more suitable for heavy work. while from the total database only 9. Wentink K. Therefore the local raw material itself. This is reflected in figure 3 displaying two distinct peaks.Grave context Ter Wal noted that TRB axes from graves are much smaller than those derived from wet contexts (Ter Wal 1996).the imported objects . It is evident that we can speak of a very selective distribution. Olausson showed.5%) while these form only 26. Frequencies of TRB dated axes per length group. is indistinguishable from the raw material used for the production of the imported axes.Neolithic Depositions in the Northern Netherlands.

The overall character of this findgroup suggests that predominantly. In figure 5 it can be seen that although a number of smaller axes was also retrieved from waterlogged places. does not explain the presence of the more extreme cases.. due to the imminent risk of end-shock. The fact that most axes found in hoards are of extreme length (>250 mm) and in mint condition might indicate that they never served as functional tools (as usage would cause an axe to wear down). A few are so big and valuable that they would not be worn at all. Fullagar & Van Gijn 2002: 18). It is striking however that the largest axes are found almost exclusively in wet contexts (finds from border contexts are included here). the majority of the axes was discarded when their decreased length/weight-ratio began to counter tool effectiveness. can be seen that many were indeed locally produced. In the Kimberley region of northwestern Australia. 4. & Quilliec B. A hypothesis therefore might be that many flint axes exceeding the length of 200 mm were not functional. These imported axes often radically changed shape. something that will be elaborated upon below in the section dealing with the results of the micro-wear analysis.1. showing that none of the axes examined longer than 218 mm displayed traces of use. Relative distribution of TRB axes from grave contexts per length group. these form only a very small proportion of the total number of axes. however. so-called Kimberly points were manufactured. but also because of practical reasons related to hafting.Hoards from the Neolithic to the Metal Ages : technical and codified practices results of the use-wear analysis. a number of axes also entered the archaeological record while being still long enough to be effective (ca. This. Some points were specially produced for exchange purposes. easily ninety per cent. 2008 ..] by far the greater number of the arm-shells. Further evidence as to the non-functional role of these large axes can be found when examining the find context from which they were derived. This can also be substantiated by the Fig. however. are of too small a size to be worn even by young boys and girls. It is however virtually impossible to distinguish between these two groups. however a number may have been worn-down remnants of imported axes. With the extremely large axes there is the question of functionality. This can partly be explained by people losing axes. as they were resharpened and repaired (Bradley & Edmonds 1993: 48). This would suggest that more practical reasons generally pertain to the discard of small flint axes. the detailed results of which are presented below. Especially with flint axes the risk of breakage (due to end-shock) increases when the axe blade is longer. .Hamon C. However. if not exclusively. worn axes accompanied the dead in their graves. Distribution of all axes from wet contexts relative to axes from other or unknown contexts. something that would undoubtedly have occurred every now and then and which is also witnessed in anthropological contexts (White & Modjeska 1978). Another example comes from Malinowski (Malinowski 1961: 88) who reports the following concerning arm-shells associated with the Kula-exchange: «[. 5. functional points (Akerman. These points could be recognized as such by their being much larger than the normal. 26% of the total number of axes were longer than 150 mm). The production of tools of extreme sizes meant for non-functional purposes is a phenomenon that is also encountered in ethnographical context. used. Fig. are only found deposited in wet contexts suggesting the need for a ritual rather than a secular explanation for discard. The extremely large axes. eds.» 34 . except once in a decade by a very important man on a very festive day.Wet context According to the explanation presented above in section 4.

Something that would further substantiate this interpretation. & A.Neolithic Depositions in the Northern Netherlands. where they were deposited either as a single object or as part of a multiple object hoard. or they could be circulated in unfinished form. would have been by not polishing them. Their size placed them apart from functional life.K. For any Neolithic person. knew upon producing it that the axe had no functional purpose. The lack of use-wear would therefore also prove that during the ‘life’ of the axe. Microwear and residue analysis Introduction The 69 objects selected came from different collections and included finds from the National Museum 35 . lacked traces of use and were often of extreme lengths all indicate that these represent a group of axes that has never been meant for usage but were specially produced of ceremonial rather that functional-related activities. This is further emphasized by some axes which are partially polished with exception of the cutting edge. After production their physical form was not altered. p. The lack of micro-wear traces. an assumption which is being substantiated by the lack of use-wear traces and the almost exclusiveness of these objects having been deposited in wet places. The fact that all of the unpolished axes came form wet locations. Apparently some axes were solely produced for non-functional purposes and ended their lives being deposited in waterlogged places. is the presence of unpolished axes in the Netherlands. but would have traveled 200-400 km before reaching the Netherlands. it would have been clear-cut that the extremely large axes would shatter upon impact. These axes traveled vast distances to reach the Netherlands without ever having been put to use. Besides extreme length a way to clearly distinguish them from Fig. There were different ways in which it could be made visually apparent that these axes should be placed apart from the functional axes. We may therefore conclude that these axes were produced for ceremonial rather than functional purposes and also circulated in this sphere. Relative distribution of unpolished TRB imported axes compared to polished imported axes. and by doing so it is placed apart from conventional. Wentink K. As can be seen in figure 6 the unpolished axes represent about 30-40% of the imported axes. a characteristic that was also recognized and respected by all people (owners?) that stood between the flint-knapper and the person/group depositing the axe 200-400 km down-the-line. leaving the cutting edge unpolished. profane tools. van Gijn . It also suggests that the flint-knapper who created the axe. They further illustrate the fact that these ceremonial axes were not tempered with. functional tools. who would have been intimately familiar with the use of flint axes. which would undoubtedly have involved exchange and transport. These unpolished axes form a well-known part of many multiple object hoards known from The Netherlands and are also often found in wet contexts with no accompanying finds. when put to functional use. 29-43 Manipulation of size can thus be regarded as a powerful strategy to emphasize the special status of an object. At the production centers in Northern Germany or Denmark axes were produced specially for ceremonial (exchange) purposes. together with the find context indicates that they should be seen in the same light as the extremely large axes. Among the imported axes there are 27 specimens which can be dated to the TRB period that are either completely unpolished or partly polished. Axes could be made to such a size they would be totally unpractical. We must also keep in mind the fact that these axes were not locally produced. When inspected for the presence of use-wear. but which is not mentioned so far. Often the latter would have been partially polished with the exception of the cutting edge emphasizing the fact that they weren’t meant for functional use. no traces of use were found on any of these objects. at no time was the axe put to a functional use. Moreover these axes ended their lives of exchange in waterlogged places. 6.

sixteen could be positively described as having been used. Moreover rounding and micro-retouch were present indicating usage. not definitely excluding the possibility that they in fact were used. & Quilliec B. Of the 69 objects 19 were grave finds coming from megalithic tombs. 7. No TRB axe larger than 215 mm appeared to have traces of use. ink. the Drents Museum in Assen and objects currently in private ownership. Some axes showed signs of patination. often exact interpretation on contact-material level was not possible. This is corroborated by the results of the use-wear analysis. which is 33.Hoards from the Neolithic to the Metal Ages : technical and codified practices of Antiquities in Leiden. The Groninger Museum in Groningen. The results of the analysis concerning SGC axes and their relation to the TRB axes are reported elsewhere (Wentink 2006). possibly also leading to the removal of potentially present prehistoric residues. This is most probably related to the fact that axes were used for all kinds of activities and not solely for the working of one contact-material. it can be seen in figure 7 that axes from nearly each group were present in the analysis. Traces of hafting could be observed in the form of friction gloss. The axes were of differing taphonomical quality. General patterns Use-traces were not equally present on all examined length groups. multiple object hoards 36 . As was already mentioned the overall character of axes coming from grave contexts could be typified as small and seemingly worn down. These differences were often accompanied by slight differences in the grinding angle and thus interpreted as being the result of secondary resharpening of the axe. eds.Hamon C. and on occasion black residue (possibly remnants of birch-tar) could be identified. The remaining two were classified as unsure. The mere handling of these objects could have obscured old traces and possibly caused new traces to develop. and glue. Of the eighteen axes. since it appeared that post depositional processes had obscured possible traces of use. 7 blades and one scraper) from several different contexts. and G2 Glimmen). the sample also contained some of the smaller axes that were analyzed for comparative purposes. Grave context Fig. 2008 . especially in grave contexts. eighteen objects were examined for the presence of use-traces. It is posited that a functional cut-off point lies around this mark. The axes came from three different passage graves (D19 Drouwen. and unpolished pieces). D5 Zeyen. Although the focus of the research lay with the large axes coming from supposed ritual contexts (single finds from waterlogged places. Moreover. Many of the axes with use-traces showed clearly developed use polish which overlies the traces of grinding. In total the database contains records of 54 TRB axes coming from grave contexts. On occasion an axe showed traces of resharpening prior to deposition. In the light of the present paper only results of the TRB axes (n=41) will be reported. With the aid of the microscope differences within the grinding-traces could be observed as indicative of the use of different grindstones. Although on some occasions a clear polish was present which indicated wood working. The sample mostly contained flint objects and predominantly axes (59 axes. When plotting the length of the TRB axes selected for micro-wear analysis relative to the total number of TRB flint axes in each length group. which could to some extent influence micro-wear analysis. Numbers of TRB flint axes selected for microwear analysis and to total number of TRB flint axes per length group. many axes had been part of museum collections. 13 were single finds (including both ‘stray finds’ and objects generally interpreted as single object hoards) and 37 objects were part of multiple object hoards. sometimes for over a hundred years.3% of the complete sample. as can be seen in table 2. 2 chisels. The fact that the largest SGC axe with use traces is 218mm long appears to substantiate this assumption. Also a variety of recent residues were encountered including such things as white paint. The micro-wear analysis embraced axes from both TRB and SGC context. nail polish. Sometimes the total lack of dirt residues indicated that the objects had been well cleaned. Of these.

& A. however each time the axe was resharpened their length decreased. van Gijn . They appeared to be caused by friction with a rather soft material. also a fair number of these were resharpened before deposition. rounding and edge-damage. Although it is impossible to tell exactly what happened. perhaps a combination of the two. Another explanation is that since these unused axes were deposited in wet contexts. The axe from tomb D5 showed minor traces of red ochre on its cutting-edge. Depositions In an attempt to isolate finds of a potential ritual character fourteen axes coming from multiple object hoards were examined for the presence of use traces.Neolithic Depositions in the Northern Netherlands. cutting-edge and higher ribs (of the unpolished axes) displayed this gloss. Some other quite interesting traces were found instead. However inside deeper negatives (caused by use) that remained untouched by the grindstone. 37 . Axes for which contextual information was available and which displayed these traces. This could have involved the object being wrapped in a certain material and being unwrapped on special occasions for display purposes. the cutting-edge predominantly displayed only very freshlooking traces of polishing. being either unpolished or much larger than average. Although the majority of the axes appear to have been used. therefore being intimately linked to the person who owned them.K. Fullagar and Van Gijn 2002). Moreover nine single finds of large axes were selected solely upon the appearance of the axe. making it as yet impossible to determine whether or not these could have been deposited axes. possibly hide or bark. All ridges. 29-43 Table 2: Micro-wear traces per context. p. With the exception of three single find axes. Also on some occasions the angle used for resharpening had left the extreme edge of the axe intact including use-traces in the form of use-polish. It is therefore not improbable that axes like these were the possessions of specific people for many years. Besides actual use-traces the majority of these axes (72. came exclusively from wet contexts. something that will be elaborated upon below. For two of these axes it was known that they came from wet contexts. all artefacts appeared to show no traces of use. Due to the overall presence of the gloss it was interpreted as having been caused by a material in which the object was wrapped in an as yet not identified material. both explanations seem plausible.2%) also displayed traces of hafting. Interesting however was the overall presence of this polish. It can be concluded that used axes predominantly accompanied the dead in the graves. In these cases. The resharpening of the axes seems to indicate that many were prepared for use to make sure that the deceased was accompanied by a sharp axe that was ready for use. use-polish could still be seen. However a very homogeneous image emerged while performing the analysis. they also played a ceremonial role prior to deposition. This is also supported by the overall worn character and minimal length of axes from grave context. On 13 of the 20 remaining axes very clear wear-traces were found. given the context and origin of these axes. an activity that is also witnessed in ethnographic context in the New Guinea Highlands (Hampton 1999) and in Northern Australia (Akerman. Moreover these small axes would have often started out as being much larger. This is not unthinkable since the axes originated some 200-400 km from Drenthe making it highly plausible that during transport the axe was wrapped in a soft material to protect it from damage. They would have been used during the clearing of fields and the construction of houses. These axes could very well have belonged to the deceased in life. of the remaining seven no detailed contextual information was known. Wentink K.

which are predominantly interpreted as being ritual in character. it is hardly possible to compare this with the distribution of the small. but presence of traces of packing/transport. Objects were individually evaluated. 2008 . in Wentink 2006). often do show traces of being wrapped in a soft material. . We know for a fact that at least a proportion of these small used axes were retrieved from wet contexts. Another interesting phenomenon encountered while examining these axes was the presence of a red residue on over 65% of the axes (table 3). .Hamon C. probably due to taphonomic reasons. as often objects cannot be dated to a specific Table 3.Hoards from the Neolithic to the Metal Ages : technical and codified practices These traces were also witnessed on some single find axes. identified as being red ochre by means of X-ray powder diffraction (Dik.Presence of red ochre residue on cutting edge. we will inevitably only see either axes coming from secure TRB contexts (mostly graves) or the large imported flint axes. This could indicate intentional deposition. The residue was not accompanied by use-traces. Spatial analysis All objects in the database have been reviewed with the above characteristics in mind. further emphasizing the fact that these axes were not meant for usage. which would have been the case if these axes were involved in some sort of contact with unprocessed ochre. for which no contextual information was recorded.Found together in a hoard with other objects that conform to either of the above stated characteristics. This might indicate that they could very well have had the same biography as those for which contextual information was present. found outside graves.Specimens longer than 218 mm (this being the largest used flint axe from the sample selected for micro-wear analysis). It can thus be concluded that the ochre most probably was applied as a pigment paste. 38 . suggesting that they would also have been deposited in wet contexts. however. . This is substantiated by the fact that the axes do not show traces of use and moreover. Late-Neolithic axes are virtually indistinguishable from TRB axes. Plotting all Neolithic axes on a map would therefore result in a palimpsest in which all potential patterns would be obscured. Although we can look for patterns within the distribution of these large TRB axes. The ochre seems predominantly located on the cutting-edge of the axes and was especially well preserved on the unpolished axes.Lack of use traces. Based on the observations described above. . since we do not know the dates of these axes. . on most axes clear traces of ochre were present all along the cutting edge.Unpolished specimens longer than 150 mm. however. to decide whether or not they should be interpreted as an intentional deposition. Different treatments per find type. the following characteristics can be presented to identify objects that are likely candidates of selective deposition: In the preceding section it has already been argued that some axes were produced solely for ceremonial purposes. eds. used axes. Another feature that distinguished these ‘ceremonial’ axes was the presence of a red pigment on the cutting edge. & Quilliec B. However. On some occasions only small fragments of red residue were encountered. they cannot be used for interpretational purposes as to TRB cultural behaviour. As mentioned before. Most objects labeled as intentionally deposited Introduction For the spatial analysis relatively few data could be used. culture. when we only plot TRB dated axes. especially when it comes to the locally produced specimens.

Bakker 1982: 114). such as at enclosures (buried in pits with pottery and flint axes at Sarup). It was already noted that many hoards with contextual information were found near the border of the peat. unpolished or coming from a multiple object hoard were automatically selected. which could be interpreted as being of a potential ceremonial character. A number of depots are located near the starting point of a stream while others are found further downstream in the valleys. sea-levels continued to rise causing the streams to become more stable and stimulated peat growth in the stream valleys (Kuijer 1991: 23). During the Atlantic. or as part of multiple object hoards (Midgley 1992. p. This would have resulted in a situation during the Middle-Neolithic in which the lower parts of the valleys would be the domain of peat growth and would be flooded during winter.K. Objects that were longer than 218mm. Kuijer 1991: 23). It is however known from Denmark that battle-axes appear in a variety of ritual contexts. including the higher grounds where they are found between the megaliths and find scatters of a potential domestic nature (hardly any settlement has been excavated). The lack of clear wood-remains indicate a fairly open landscape with only few trees. Natural landscape of depositions Fig. resulting in the formation of large strings of small bogs and fens (Spek 2004: 203. 101). Spatial Distribution of TRB depots on the Drenthe Plateau. These were present in all zones. In the Weichselian these valleys were deepened and widened. a clear pattern emerged. The same applies to many of the TRB battle axes that have gone unmentioned so far. When the 36 depots that were selected as likely candidates of selective deposition were spatially plotted. Depositions only seem to occur here in the direct vicinity of an intersecting stream. Objects that scored positive on the presence of ochre or presence of traces of packing/ transport were only selected if they also scored on any of the other characteristics. Neolithic people did not 39 . Axes dated to the TRB but which were not included in the selection. The obvious lack of selected axes in the most extensive raised bogs such as the Bourtanger Bog. 86. must therefore be noted. Also a high proportion of TRB axes between 150 mm and 218 mm appeared to be located in the stream valleys possibly suggesting that these also concern deposited objects. & A. Due to the Holocene rise in groundwater-levels (as a result of rising sea-levels) these obstructions eventually eroded and streams reemerged. Tilley 1996. Within these valleys the depots are predominantly found at the border of the peat. showed a far more diverse distribution pattern. 8. in contrast to the higher grounds on which a dense forest was present (Spek 2004: 209. This group therefore would be an interesting subject for further research. The stream valleys in which the selected axes were found formed predominantly in the Pleistocene. Wentink K. 8). van Gijn . when at the end of the Saalian ice-age streams of meltwater eroded their way through the newly formed till plateau (Spek 2004: 203). Skaarup 1990. Virtually no contextual research has been carried out for this group in the Netherlands. 245. When examining the spatial distribution it is striking to see that virtually all selected axes are located in stream-valleys that would have been filled up with peat (see fig. which would have been the most practical position since then the person(s) depositing the axes would not have to enter the potentially dangerous peat zone. This method resulted in the selection of 55 axes belonging to 36 depots. However at the end of the Weichselian the valleys were blocked by large deposits of cover-sands.Neolithic Depositions in the Northern Netherlands. however in Drenthe peat-growth was common in many different places. Although archaeologists usually focus upon the peat itself in their explanations. 29-43 conformed to more than one of the above mentioned characteristics. but were also often found in stream valleys.

historical sources indicate the existence of specific names to describe such places from the Middle Ages onwards (Spek 2004: 206). Together with the observation that part of the hoards are located at the beginning of streams we might envisage that people were depositing items at places where water would emerge from the ground. which would have caused high evaporation rates. This is in contrast to many of the find-scatters. It is known that in historical times at these places water would seep from the ground. This also appears to be the case in the Netherlands as clearly many locations can be found where a depot was located close to a megalithic tomb. By no means should this be interpreted as settlements being located near graves. Although this suggests a link between the two it also indicates a separation since no depot was found less than 600 m from a grave. For Denmark it was noted that half of the hoards were found within close proximity (500-1500 m) to megalithic graves (Midgley 1992: 282). however. Also groundwater levels would have been influenced by the dense forest on the higher grounds. Many flat-graves will not have been discovered. comm. Spatial distribution of TRB sites on the Drenthe Plateau. At this point it is unfortunately impossible to prove such a scenario. eds. The point being made however is that we should not solely focus on the peat as being of prime importance. non-permeable layers of sediment (till. possibly remnants of settlements. Since the first is generally lacking the latter is impossible to gather.Hoards from the Neolithic to the Metal Ages : technical and codified practices necessarily do the same thing. due to their naturally obscured nature. It could very well be possible that hoards that do not conform to these observations are instead located in proximity to undiscovered flat-graves.). it is clearly stated that finds were retrieved from the sand. As many depots are located at the transitional point in these valleys between the lower peat and the higher sands.Hamon C. These places should have been recognized by people from obvious differences in local vegetation (oak/birch forest) (Spek 2004: 206). near the peat and not from the peat itself. At some locations on the higher grounds in the valleys. depicted in figure 9. For some hoards for which detailed contextual information is present. These places would not have been clear wells. Many find-scatters are found isolated in the landscape with no indication as to the presence of nearby graves. which are located within 500 m of a grave. & Quilliec B. This indicates that people did recognize such places and could attribute special meaning to them. It is impossible to predict exactly where these places would have been located in the Neolithic (due to highly variable groundwater-levels and local geology). 9. pers. loam) were present beneath the sand. thus lowering groundwater levels (Spek. but also proximity to flat-graves can be noted. Cultural landscape of depositions The spatial distribution of the 36 selected depots conformed very well to the overall distribution of TRB sites. 2008 . About half of the depots (53%) could be found within a range of 600-1900 m from the nearest grave. that most of the hoards in the stream valleys are located on a soil type that would be expected in the above scenario. whereas all depots were kept well outside this ran40 . the latter may potentially be equally important. Fig. It is striking. were not necessarily deposited in the peat but might well have been engulfed by it during later times. Furthermore the rise in groundwater-levels caused the peat to grow and to cover areas that in the Neolithic would have been sand. It does however imply that it was apparently not considered problematic to locate a settlement within 500 m from a tomb (or vice versa. Therefore axes recovered during peat digging or during other activities on the land that would formerly have been covered with peat. however. assuming these find-scatters represent settlements). as it would require the exact find location of each depot and also detailed geological information of that find spot.

Een depot van vuurstenen bijlen bij Reest.. The places selected for deposition at the transition from marshes to the higher. located near the cutting-edge of the axe.A. We can therefore conclude here with the observations that flint axes played an important role in TRB cosmology/ideology. FULLAGAR R. BAKKER J. but at the same time also as places binding social groups. van Gijn . Appropriate places where depositions took place were selected in close range of places of burial and habitation. This separation was further emphasized by red pigments (ochre). which will be reported upon in future publications. This would provide information about the homogeneity as well the local nuances of this ritual. On the one hand these stream-valleys will have been perceived as natural boundaries between social groups as well as boundaries between people and supernatural entities (Fontijn 2002: 265). 1979. TRB Settlement Patterns on the Dutch Sandy Soils. thus connecting social groups. 1960. PhD thesis Amsterdam. It is unfortunately impossible to tell the exact role these axes played during ceremonies and rituals. 1993. dry grounds are therefore of a liminal nature. References ACHTEROP S.K.). Weapons and wunan: production. the spatial patterns seem to suggest the existence of a well-defined ritual.Neolithic Depositions in the Northern Netherlands. 2002. functional objects.H. Together with the very consistent micro-wear patterns involving traces of packing or wrapping and the residue of red ochre. however at the same time a certain distance was maintained between the two. On the other hand. They were specially produced and were exchanged over vast distances. (eds. Australian Aboriginal Studies 1: 13-42. BRADLEY R. On a physical level these places can therefore be perceived as clear-cut divisions between the higher habitable grounds and the natural waterlogged stream valleys. p. Nieuwe Drentse Volksalmanak: 17989. Interpretation of the role these axes played within their cultural context will be the aim of further study. These depositions occur over the entire Drenthe Plateau. & VAN GIJN A. function and exchange of Kimberley points. at the same time an appropriate distance between the two was observed. that these objects had to be placed apart from profane. Analecta Praehistorica Leidensia 15: 87-124. 41 . 1982. Although the objects deposited are of a non-local origin the practice of depositing however is very much a local affair. These axes would often have been visually set apart from other axes by their size or by being unpolished. seems to indicate that this ritual is closely intertwined with TRB cosmology/ideology. Wentink K. the higher grounds were densely forested. & EDMONDS M. AKERMAN K. On a cultural level these places can both be perceived as boundaries separating social groups. During transport they would have been packed in soft material to protect them from damage. Conclusion The present paper set out to report the preliminary results of the research currently being carried out. Depots occurred relatively near graves. Habitation and tomb-construction primarily took place on the higher grounds of the Drenthe Plateau.A. Cambridge: Cambridge University Press... Their biographies however ended when they were deposited near the edge of the peat in one of the numerous stream valleys present on the Drenthe Plateau. Production and exchange in Neolithic Britain. BAKKER J. Although these places were located near the places of burial and habitation. indicating that this ritual was widely adopted and performed by the TRB people. The many stream valleys would therefore have played an important role in water transport. The TRB west group : studies in the chronology and geography of the makers of hunebeds and Tiefstich pottery. The axes that were found in depots were probably used in rituals or ceremonies. By doing so it was clearly emphasized again and again. & A. Interpreting the axe trade. They were kept apart from functional tools at all stages of their lives. The fact that these axes were specially produced for ceremonial or ritual practices combined with the fact that they are found deposited in wet places over so vast an area of Northern Europe.. 29-43 ge. It would therefore be fascinating to extent this analysis to the deposited axes from Germany and Denmark.

An Evaluation of Raw Materials Based on Experiment. Ceci n’est pas une hache. PhD thesis Nijmegen. KUIJER P. 1990. (ed. Votive Offerings and Social Structure in Early Neolithic Farmer Society of Denmark. Palaeohistoria 37/38:127-58. 2005. 2004.. London: Routledge. Een onderzoek naar de depositie van vuurstenen bijlen..). Scripta Minora WHITE J. SCHUHMACHER K.S.H..W. HOGESTIJN J.Drenthe..).S. TER WAL A. Barkcloth and the Dynamics of Identity in Papua New Guinea. 1992. Burials. & Quilliec B.Hamon C. VAN GIJN A. Early prehistoric societies in southern Scandinavia. Poznan (Material des Internationalen Symposiums Dymaczewo 20-24 September 1988). WENTINK K.. STOUT D. An ethnography of the Neolithic.H.J. 277-80. PhD thesis Leiden (Analecta Praehistorica Leidensia 22).W. Fokkens H.. 2300-600 B.).. C. Users. Amsterdam: Bert Bakker.C. HAMPTON O.P. Neolithische Depotfunde im westlichen Deutschland. (eds. The Wear and Tear of Flint: Principles of Functional Analysis to Dutch Neolithic Assemblages. Prehistoric Disc Wheels in the Netherlands..). Engendering objects. 1983. 1964. 2002. (Regiae Societatis Humaniorum Litterarum Lundensis) PLEYTE W. Amersfoort (ROB Rapportage Archeologische Monumentenzorg 79). VAN DER WAALS J. an account of native enterprise and adventure in the archipelagoes of Melanesian New Guinea. 73-93. SPEK T. LOUWE KOOIJMANS L. PhD thesis Leiden (Analecta Praehistorica Leidensia 33/34). & PEETERS J. Nederland in de Prehistorie.R. 2001.. 2002. 2005. TRB Culture – The First Farmers of the North European Plain.. Flint and Groundstone Axes in the Scandinavian Neolithic. Neue Forschungen und Hypothesen. 1882. 1999. (ed. PhD thesis Groningen.. MALINOWSKI B. The Flourish and Demise of an Old Technology. n° 2). (eds. & Van Gijn A. 2001. OLAUSSON D.. 2006.. Current Anthropology 43/5: 693-722... Finders. Texas (Texas A&M University Anthropology Series. Wageningen (Stiboka). FONTIJN D. Brill.. Leiden (Sidestone Press). 1996. Bodemkaart van Nederland 1:50 000.L. Nederlandsche oudheden van de vroegste tijden tot op Karel den Groote . in Louwe Kooijmans L. VAN GIJN A. eds. objects and ‘natural’ places in the Bronze Age of the Southern Netherlands. in prep. 42 .Sacred and Profane Uses of Stone among the Dani. MIDGLEY M. Die Trichterbecherkultur.Hoards from the Neolithic to the Metal Ages : technical and codified practices BEUKER J. De mesolithische en vroeg-neolithische vindplaats Hoge Vaart-A27 (Flevoland) – Deel 20 Op de grens van land en water: jagers-vissers-verzamelaars in een verdrinkend landschap.W. Sacrificial Landscapes Cultural biographies of persons. Culture of Stone .. Chr).. 1978. SKAARUP J. Van den Broeke P. Utrecht: Matrijs. Losers: The Use Axe Blades Make of the Duna. Leiden: E. Cambridge: Cambridge University Press. Argonauts of the Western Pacific. Toelichting bij kaartblad 12 West Assen.M.D. Amersfoort (ROB Rapportage Archeologische Monumentenzorg 88).P..K. 1991. Archeologie in de Betuweroute: Hardinxveld-Giessendam De Bruin.C.P. Een historisch-geografische studie. Acquirers. 1996. Teil I. & MODJESKA N. HERMKENS A. Prahistorische Zeitschrift 6: 29-56. Mankind 11: 276-87. 1914. Neolithic Deposition in the Northern Netherlands.. Een kampplaats uit het Laat-Mesolithicum en het begin van de Swifterbant-cultuur (5500-4450 v. Edinburgh: Edinburgh University Press. TILLEY C.L. Skill and Cognition in Stone Tool Production . Aanvoer uit alle windstreken Stenen bijlen in Noord-Nederland. The Meaning of Flint for Neolithic and Bronze Age Societies in the Netherlands. Research Master thesis. 1990.. in Jankowska D. 2008 . Het Drentse esdorpen-landschap. 1961(1922).An Ethnographic Case Study from Irian Jaya.

K. Quentin 43 . Benoit Mater and Vincent van Vilsteren (Drents Museum). Groningen University). which was gratefully incorporated. notably Jaap Beuker. Faculty of Archaeology PB 9515. & A. Harry Fokkens (Fac. Special thanks also to MA & PhD students Stijn Arnoldussen. Leiden University). p. We would also like to thank the many people who provided data and answered numerous questions.gijn@arch. 2300 RA Leiden The Netherlands k. 29-43 Acknowledgments Several people of the different museums we visited through the years have shown great patience and helpfulness.van.wentink@arch. Leiden University) and Sake Jager (State Archaeological Service) both gave access to their unpublished data concerning axes.l.leidenuniv. van Gijn . Alice Samson and Corné van Woerdekom who have critically read this text. Ernst Taayke (Archaeological Depot of the Northern Netherlands) and Leo Verhart (National Museum of Antiquities).Neolithic Depositions in the Northern Netherlands. Contact Karsten Wentink Annelou van Gijn Laboratory for Artefact Studies. of Archaeology. Peter Schut (State Archaeological Service) and Theo Spek (State Archaeological Service). Leendert Louwe Kooijmans (Fac. Wentink K.leidenuniv. Wijnand van der Sanden (Province of Drenthe).nl a. Jaap Beuker (Drents Museum). Piet Kooi (Department of Archaeology. the first author would like to thank Caroline Hamon and Bénédicte Quilliec who invited him to speak in Cork. Erik van Rossenberg. of Archaeology. Finally. Leiden University). of Archaeology. David Fontijn (Fac.